1. Field of Invention
This invention relates to a method and apparatus for generating enhanced thumbnails.
2. Description of Related Art
Computer users spend a significant amount of time examining collections of documents, such as documents retrieved by a search engine from the Internet. The user must page through lists of documents, briefly evaluating each for possible relevance to a particular information need. Improving the efficiency of this tedious process would directly benefit the end-user and, by improving end-user satisfaction, would indirectly benefit parties such as the search engine vendor.
For instance, the Internet search engine can increase user efficiency by (1) returning higher-quality document lists (e.g., through better index coverage and ranking algorithms) or by (2) providing information that allows the user to evaluate the results more quickly and accurately. Search engine vendors attack both problems. The standard practice with regard to approach (2) is to provide brief textual summaries of the Web documents. In recent years, in addition to textual summaries of documents, it has been suggested that graphical summaries of the documents, such as thumbnail images, can greatly increase the efficiency by which end-users process search engine result sets.
Each of these approaches has advantages and disadvantages. For example, text summaries are terse but are verbal rather than visual. They require little storage space and can therefore be downloaded quickly. Additionally, text summaries often contain a great deal of valuable information about each document. For example, search engines commonly provide the document's URL, header, size, and a few phrases or sentences that either summarize the document or emphasize some of the search keywords. On the other hand, text summaries do not provide much information about the page layout or any image contained in the page. Furthermore, the user must read the text summary, which is time consuming and tiring.
In contrast, graphical summaries do provide information about the layout, genre, and style of the page. If the user has previously seen the page, or one like it, the visual representation may aid in recognizing or classifying it. This becomes even more compelling in view of the fact that the human visual system can process images more quickly than text. Graphical information can speed many tasks tremendously. However, thumbnails typically require more storage space than text summaries, and therefore, they generally download more slowly than text summaries. Further, textual content in plain thumbnails is less accessible than that in text summaries, as it is difficult to read and is not conveniently summarized.
Previous work includes several different designs for thumbnails. A number of programs generate plain thumbnails. These include many graphical editors, recent versions of Microsoft Windows, and systems described by Hightower et al. (Hightower, R., Ring, L., Helfman, J., Bederson, B., & Hollan, J. (1998), Graphical Multiscale Web Histories: A Study of PadPrints, Proceedings of Hypertext ″98, 58–65), among others. Thumbnails of Ayers et al. (Ayers, E., & Stasko, J. (1995), Using Graphic History in Browsing the World Wide Web, Proceedings of the 4th International World Wide Web Conference) are similar to plain thumbnails, showing a reduced view of the upper left corner of a document.
Other programs generate more complex thumbnails. Cockburn et al. (Cockburn, A., Greenberg, S., McKenzie, B., Jasonsmith, M., & Kaasten, S. (1999). WebView: A Graphical Aid for Revisiting Web Pages, Proceedings of OZCHI″99, Australian Conference on Human Computer Interaction) generate thumbnails that show reduced images plus dogears that indicate bookmarked and frequently visited pages. Helfman (Helfman, J. I. (1999), Mandala: An Architecture for Using Images to Access and Organize Web Information, Proceedings of Visual ″99, Visual Information and Information Systems, Third International Conference, 163–170) selects representative images from a document and creates reduced scale images of these to serve as a thumbnail for that document. Wynblatt et al. (Web Page Caricatures: Multimedia Summaries for WWW Documents; IEEE; 1998) produce Web page caricatures. These caricatures contain select features of a page, often rendered in an abstract form, such as the header, a representative image, the number of images, an abstract, etc. However, these caricatures do not preserve layout and lack some of the visual information that might be naturally available in a reduced scale image of the page. For example, rather than having the user judge link density, or the number of links on a Web page from an image of the page, this link density is represented by the background color of the caricature.